Canon 40D Image Quality? What's a Good Alterntive Camera Body?

[John Knight]

I've had a Canon 40D since December 2007 which I've used for bird and other wild life photography. I've recently come to the conclusion that the image quality is not as good as the Canon 5D. I've checked several on-line discussion threads, and it seems that others have reached a similar conclusion about the 40D.

When I've asked other photographers about the Canon 1D Mark III as an alternative, I've been advised to wait and see what Canon does with the various fixes with image tracking and quality. The camera store that I use locally is also telling me not to consider the 1D Mark III, and to perhaps look for a good 1D Mark II.

Are others finding the same shortfall is image quality with the 40D? Is the 1D Mark III worth considering, and why, or why not?

In the meantime, I'm back to using my 5D, but missing 6.5 fps and 1.3X factor on the sensor of the 40D ... but happier with overall image quality.

[Axel Hildebrandt]

Could you elaborate what you mean by better/worse image quality? The 40D is a 1.6x crop sensor and the pixel density is high, which might lead to higher ISO noise than in the 5D.

[Robert O'Toole]

The 40D (and the 5D) is a great camera and has a huge potential to make great images. The files are clean and the camera has some great features. Whats not to like?
I would always take the 40D over the 5D for bird photography and especially BIF.

The only problem I can see with the 40D, and the 5D, is that the AF specs are very underpowered compared to the competition out there. But thats another issue :)

Robert

[John Knight]

I have found that when I examine my images from the 40D in Bridge using the loupe, they appear to have a 'misty, soft appearance' when I examine the eyes, and they seem to be a little less sharp than those from the 5D, both using the same lenses, typically a 300 mm f2.8 with a 1.4X tele-converter with a Gitzo G3540XLS tripod, RRS ballhead and a cable release. There are a lot of good aspects in the 40D (1.6X cropping factor, 6.5 fps, live-view, etc.), but I was a little surprised to find that my images weren't quite as sharp as those from the 5D. The images that I've been comparing are static bird shots, not birds in flight. For the latter, I've been trying take advantage of the higher fps in the 40D.

Perhaps I need to think of the 40D for use primarily for BIF (dynamic shots), and the 5D for birds at rest, or moving slowly.

John

[Robert O'Toole]

That makes sense John, for years and years I used a EOS-1D (MKI,MKII,MKII N) and a 1DS (MKI and II) the same way. If the bird was moving I used my 1D and for everything else my 1DS. I always had a 20D in my pocket for back up or when I need the reach.

Try more sharpening in your 40D settings, they are sharper by default than earlier Canons, but you might like to see more when you view the Jpegs (you really view an embedded JPEG when you view a CR2 file)

Robert

[Brian Wong]

...(you really view an embedded JPEG when you view a CR2 file)

Robert

Hi Robert!

How are you?

I believe most people are really not aware how profound this statement can be. Please correct me if I am wrong, but my observation is that many, many applications will display the embedded JPEG from the CR2 RAW file ... (mainly for speed & memory management purposes).

[John Knight]

That's a really good point regarding Bridge. I've into the trap of reviewing my images only in RAW since I acquired Bridge, and recently I haven't been doing any printing of my images.

If I were to consider upgrading either a camera body or a lens, which should be my priority? What kind of improvements might I expect if I were to upgrade to a 1D Mark III? Does the 1D Mark III really have the problems that you read and hear about? Or would upgrading to a 500mm f4 be a better choice, rather than continuing with my 300mm f2.8 + 1.4X?

I've only recently become a member of Birdphotographers.net, and I'm just getting into bird photography. It's great to be able to ask questions like these. Thank you for your help and advice ... any and all is welcome.

John

[Robert O'Toole]

Hi Robert!

How are you?

I believe most people are really not aware how profound this statement can be. Please correct me if I am wrong, but my observation is that many, many applications will display the embedded JPEG from the CR2 RAW file ... (mainly for speed & memory management purposes).

Hi Brian!

Yes, you are right. Most people dont realize that when you view the image on the back of the camera you are looking at the JPEG thumbnail. When you download the file to your computer to examine the image on a viewing program you are looking at the embedded JPEG. The RAW file is not really designed to be viewed. That is the reason there is a JPEG embedded in the file.

Robert

[Robert O'Toole]

If I were to consider upgrading either a camera body or a lens, which should be my priority?

I would upgrade the glass first and wait for a EOS-1D MKIV if was in your situation. The 40D is a very capable camera and the money would be better spent on a new lens, unless your budget can handle both.

I decided to buy a Nikon D300 in Nov 2007 to take advantage of both systems.

What kind of improvements might I expect if I were to upgrade to a 1D Mark III? Does the 1D Mark III really have the problems that you read and hear about?

This subject had been covered in depth here and on other sites on the internet.

After 5 years of using Canon pro bodies I have started using Nikon because of the MKIII customer service and design issues. I know of 6 friends, and more than a few acquaintances that have done the same. People do shoot with MKIII and do well but since the camera has been released there are now other excellent options from Nikon that really offer much newer technology and features not available on the Canons.

Or would upgrading to a 500mm f4 be a better choice, rather than continuing with my 300mm f2.8 + 1.4X?

I would recommend the 500/4 or 600/4 for birds.

Robert

[John Knight]

Thanks Robert. I appreciate the advice and feedback.

In speaking with a long time pro this afternoon, he suggested that I do some rigorous image tests with my 5D and 40D to further verify the nature of any differences between the two camera bodies. He wondered if my 40D might need some adjustment by Canon if there is a significant difference in sharpness while using the same lens. I'll check this out.

John

[Robert O'Toole]

Good plan, a back=front focus test on a 45 deg ruler would tell you if the softness is AF related. It could very well be.

Keep us updated, I am interested to see what the outcome is.

Email or PM if you need any help.

Robert

[William Malacarne]

John

I am sorry I didn't see where you said what type of file you are looking at for sharpness. Is it a RAW or a jpg from the camera?

Bill

[Brian Wong]

I have found that when I examine my images from the 40D in Bridge using the loupe, they appear to have a 'misty, soft appearance' when I examine the eyes, and they seem to be a little less sharp than those from the 5D, both using the same lenses...
John

Hi John!

I also think that it is difficult to actually compare varoius CR2 RAW file between cameras. As far as I know, the files all need to be sharpened to some degree. This is primarily due to the anti-alias filter that is placed in front of most camera sensor to counteract the configuration of Bayer sensor designs. I do not recall where I read this ... but it is my impression that the camera companies are coming out with stronger and stronger anti-alias filters with each newer model. Maybe they are thinking that the combination of strong filters along with strong user sharpening is the best combination for a good image???:confused: Or maybe they are trying to prevent us from the dreaded "oversharpening"???:confused:

Just some thoughts, as with you ... just trying to learn and understand more what is going on inside those machines.

[William Malacarne]

I have found that my 40D photos straight from the camera(RAW) are not as sharp as the ones from my 20D or 1DII N. But when they are processed they stand right up with the others without any problem.

Bill

[David Kennedy]

I borrowed a 40D from CPS back at the beginning of March and compared it to the 5D. It reminded me a lot of the 10D and 20D, except that the files were waaaaaaaaaay smoother (thank you, 14-bit!). But the 40D's files reminded me of the earlier bodies because the pixel density is more akin to those other APS-C cameras than the 5D. The 5D simply resolves more detail, but that doesn't make the 40D unsharp. You've just been spoiled by the large files made on a large sensor (with photo sites spread out much more so than on the APS-C sized sensor).

[edward saltau]

Hi Brian!

Yes, you are right. Most people dont realize that when you view the image on the back of the camera you are looking at the JPEG thumbnail. When you download the file to your computer to examine the image on a viewing program you are looking at the embedded JPEG. The RAW file is not really designed to be viewed. That is the reason there is a JPEG embedded in the file.

Robert

Hi robert,
I know breezebrowser uses the embedded jpegs. This is one of the main reasons why i like BB because it makes it so fast and easy to use. But recently I haven't had access to BB and have had to cull my shots in lightroom on my mates mac. Does lightroom use the embedded jpeg? I'm guessing it doesn't because it seems really slow compared to BB. should we cull based on the jpeg or should we cull based on the raw? hhmmmm

[John Chardine]

It should be no surprise that the 40D IQ is not equal to the 5D as the sensor characteristics of the two are quite different. The 5D is a top performer as shown in the various analyses by R.N. Clark who is BPN member I think:
http://www.clarkvision.com/imagedeta...mance.summary/

BTW I'm a 40D owner and like the speed but am under no illusions that it is equal to a full-frame sensor body.

[Robert O'Toole]

Hi robert,
I know breezebrowser uses the embedded jpegs. This is one of the main reasons why i like BB because it makes it so fast and easy to use. But recently I haven't had access to BB and have had to cull my shots in lightroom on my mates mac. Does lightroom use the embedded jpeg? I'm guessing it doesn't because it seems really slow compared to BB. should we cull based on the jpeg or should we cull based on the raw? hhmmmm

Yes, LR renders new thumbnails and does not use the embedded thumbnails . These previews are based on your conversion settings and LR stores them in a .lrdata file. These files can sometimes grow to Gb size since LR does not delete these thumbnails even if you delete the RAW files. I have heard of 15GB .lrdata files. You can change the prefs to throw the files away though. Here is an adobe link that talks about it in case you want to know more: Adobe .lrdata link


Robert

[Robert O'Toole]

Hi robert,
I know breezebrowser uses the embedded jpegs. This is one of the main reasons why i like BB because it makes it so fast and easy to use. But recently I haven't had access to BB and have had to cull my shots in lightroom on my mates mac. Does lightroom use the embedded jpeg? I'm guessing it doesn't because it seems really slow compared to BB. should we cull based on the jpeg or should we cull based on the raw? hhmmmm

Sorry Edward I missed one of your points.

I use BB or Bridge to view my RAW files. BB is so much faster and they also tend to look better with the extra sharpening and contrast enhancement in the Hi Quality setting. Otherwise the files tend to look flat in bridge since my ACR defaults are custom and not default.
So I think it is better to view the embedded Jpegs since this speeds the process up and the images will look exactly as they did on the LCD preview in camera as shot.

Robert

[John Knight]

I did the back-front focus test on both the 40D and 5D this morning, and both look normal and about the same in this regard. So, it's good that this is not the problem.

I took some pictures of a robin recently on both the 40D and 5D, and again discovered that the 5D images were sharper. I suspect that the issue may be 'operator error', me, with respect to where the focus point(s) intersect what I'm photographing. I need to pay more attention to this aspect, and perhaps will find that this is what is creating the sharpness differences.

And yes, the images do look better from both cameras when displayed as JPEGs or TIFFs.

Thanks everyone for your input and feedback ... I'm learning.

John

[Robert O'Toole]

I also think that it is difficult to actually compare varoius CR2 RAW file between cameras. As far as I know, the files all need to be sharpened to some degree. This is primarily due to the anti-alias filter that is placed in front of most camera sensor to counteract the configuration of Bayer sensor designs. Just some thoughts, as with you ... just trying to learn and understand more what is going on inside those machines.

Some companies actually offer Anti-alias or Low-pass filter removal services. The results are supposedly stunning.

Also AFAIK the old Kodak FF camera and the leica digital back did not use AA or LP filters and were sharp as heck, but also had serious moire problems.

Some high $ medium format digital cameras offer removable AA/LP filters that you can set for the subject requirements.

Wouldnt it be great if the D3X or the 1DMKIV had this replaceable AA/LP filter feature. Well it will never happen but its still ok to dream right?

Robert
p.s. LP filter removal surgery on a D2H : http://nikongear.com/smf/index.php?topic=7788.0

[David Kennedy]

The original Canon 1D also lacked an AA filter and was also ridiculously sharp. And it had 1/500 sec. flash sync. . .

[Robert O'Toole]

Oh I remember that camera, I remember paying over $5000 each fof 4.1 mp @ 8 fps, and the tiny 2400 pixel file sizes. I remember 200-250 shots per battery, and I remember how much those 5 extra $150 batteries weighed......:(

But i dont remember the missing AA/LP fliter.....I thought all the C dSLRS since the D30 had them....

Are you sure David?

Robert

[Robert O'Toole]

Looks like Phil askey doesnt support the non-AA filter theory either :)

http://www.dpreview.com/reviews/CanonEOS1D/page19.asp

Moiré
Canon have noted previously that they are using a 'less powerful' (eg. less intrusive) anti-alias filter in front of the sensor. This improves sharpness but does increase the amount of work which has be done by the cameras internal processing algorithms to remove any artifacts created by high frequency detail (moiré patterns). On the whole this appears to work very well, the camera delivers great resolution. But there were occasions (a very small % of my shots) where the right type of detail at the right frequency was able to bypass Canon's trickery and appear in the final image as a moiré pattern

So maybe its a less powerful AA/LP than the D30?
I vaguely remember an interview the Canon 1D chief engineer and he said that no one should ever touch the low pass filter on the 1D with anything. I have no idea where I read that though....
Robert

[Brian Wong]

Looks like Phil askey doesnt support the non-AA filter theory either :)

http://www.dpreview.com/reviews/CanonEOS1D/page19.asp

Moiré
Canon have noted previously that they are using a 'less powerful' (eg. less intrusive) anti-alias filter in front of the sensor. This improves sharpness but does increase the amount of work which has be done by the cameras internal processing algorithms to remove any artifacts created by high frequency detail (moiré patterns). On the whole this appears to work very well, the camera delivers great resolution. But there were occasions (a very small % of my shots) where the right type of detail at the right frequency was able to bypass Canon's trickery and appear in the final image as a moiré pattern

So maybe its a less powerful AA/LP than the D30?
I vaguely remember an interview the Canon 1D chief engineer and he said that no one should ever touch the low pass filter on the 1D with anything. I have no idea where I read that though....
Robert

While this is not directly related to the original 40D question, in this month Tech Tips (May) issue, Chuck Westfall, Director of Media & Customer Relationship for Canon USA, touches on the issue of sharpening and the anti-alias filter of the various Canon bodies through the years. He also addresses the "Picture Style" algorithm currently used by Canon in the apparent effort to unify, or standardize the latest models. It does appear that different strength anti-alias filters are used with different model sensors. He does state that all native sensors are "sharp", which makes sense to me. It is my impression that the various comparison efforts may have more to do with all the various manipulation (physical hardware, and software processing, and cost factors) that the camera manufacture install into their bodies, which gives us the end result that we work with (this includes both "JPEG" and "RAW").

The article can be found here:

http://digitaljournalist.org/issue0805/tech-tips.html